The AI in Business Podcast Why Manufacturing's Most Valuable Data Isn't in Any System — with Anand Gnanamoorthy of Ingersoll Rand
29 snips
May 13, 2026 
Anand Gnanamoorthy, Director of Corporate Strategy and AI at Ingersoll Rand, works on digitizing tribal knowledge and applying AI to frontline workflows. He discusses the urgency of capturing retiring workers' knowledge. He highlights messy unstructured archives as the biggest untapped asset. He urges anchoring AI to workers, keeping projects perpetually in pilot, and letting AI handle messy data.
AI Snips
Chapters
Transcript
Episode notes
Define Clear Human Versus AI Decision Boundaries
- Decide which decisions AI should make and which humans must keep; treat AI as an insights engine, not an autonomous decision-maker for sensitive choices.
- Anand warns pilots fail when teams don't define the human-versus-AI decision boundary beforehand.
Three Layers Where Manufacturing Knowledge Hides
- Most manufacturing knowledge lives in three layers: structured operational data, decades of unstructured archives, and tribal knowledge in employees' heads.
- Anand says dusty emails, letters, word files and drives across 40–50 years often mirror an expert's knowledge and are the biggest untapped source of value.
Let AI Clean Messy Archives Instead Of Manual Scrubbing
- Do not over-clean archival data before using AI because modern models can handle messy, duplicate-heavy records and surface the latest relevant versions.
- Anand suggests letting AI deduplicate and find authoritative documents instead of manual mass cleaning.